Discovery of a very X-ray luminous galaxy cluster at z=0.89 in the WARPS survey

a r X i v :a s t r o -p h /0012175v 1 8 D e c 2000

Draft version February 1,2008

Preprint typeset using L A T E X style emulateapj v.04/03/99

DISCOVERY OF A VERY X-RAY LUMINOUS GALAXY CLUSTER AT Z =0.89IN THE WARPS

SURVEY

H.Ebeling 1,8,L.R.Jones 2,8,B.W.F airley 2,E.Perlman 3,4,C.Scharf 5,D.Horner 6,7

accepted for publication in ApJL

ABSTRACT

We report the discovery of the galaxy cluster ClJ1226.9+3332in the Wide Angle ROSAT Pointed

Survey (WARPS).At z =0.888and L X =1.1×1045h ?250erg s

?1

(0.5?2.0keV)ClJ1226.9+3332is the most distant X-ray luminous cluster currently known.The mere existence of this system represents a huge problem for ?0=1world models.

At the modest (o?-axis)resolution of the ROSAT PSPC observation in which the system was detected,ClJ1226.9+3332appears relaxed;an o?-axis HRI observation con?rms this impression and rules out signi?cant contamination from point sources.However,in moderately deep optical images (R and I band)the cluster exhibits signs of substructure in its apparent galaxy distribution.A ?rst crude estimate of the velocity dispersion of the cluster galaxies based on six redshifts yields a high value of 1650km s ?1,indicative of a very massive cluster and/or the presence of substructure along the line of sight.While a more accurate assessment of the dynamical state of this system requires much better data at both optical and X-ray wavelengths,the high mass of the cluster has already been unambiguously con?rmed by a very strong detection of the Sunyaev-Zel’dovich e?ect in its direction (Joy et al.2001).

Using ClJ1226.9+3332and ClJ0152.7–1357(z =0.835),the second-most distant X-ray luminous cluster currently known and also a WARPS discovery,we obtain a ?rst estimate of the cluster X-ray

luminosity function at 0.85×1044h ?250erg s

?1

(0.5?2.0keV).Using the best currently available data,we ?nd the comoving space density of very distant,massive clusters to be in excellent agreement with the value measured locally (z <0.3),and conclude that negative evolution is not required at these luminosities out to z ～1.Our ?ndings are in con?ict with earlier claims of highly signi?cant (>3σ)negative evolution already at 0.3

Subject headings:galaxies:clusters:general —galaxies:clusters:individual (ClJ1226.9+3332,

ClJ0152.7–1357)—cosmology:observations —X-rays:general

1.INTRODUCTION

Measurements of the abundance of clusters of galaxies as a function of redshift allow a number of physical and cosmological parameters of structure formation models to be constrained (e.g.Oukbir &Blanchard,1997;Eke et al.,1998).The tightest constraints are obtained from ob-servations of the most massive and most distant clusters which are extremely rare in all models of cluster formation.For instance,the predicted space density of galaxy clus-ters with intra-cluster gas temperatures of about 8×108K (k T ～7keV)at z ～1is two orders of magnitudes higher in an open universe with ?0=0.3than in a ?at universe with ?0=1(Viana &Liddle,1996).

Among the many e?orts to compile statistically rele-vant samples of these rare systems,X-ray ?ux limited sur-veys carry particular appeal.Galaxy clusters are bright X-ray sources that can be detected out to high redshift.Moreover,X-ray emission from clusters originates from gas trapped and heated in deep gravitational potential wells.X-ray surveys thus naturally select three-dimensionally bound systems and are almost una?ected by projection ef-fects.Finally,X-ray ?ux limited surveys greatly facilitate the measurement of comoving space densities since e?ec-tive search volumes are easily computable from the sur-veys’selection functions which are usually a simple func-tion of X-ray ?ux and,for distant clusters,almost inde-pendent of X-ray source extent.

While one would ideally want to study clusters that are both distant and massive,the rarity of these objects cur-rently forces the observer to give priority to one of the two

1Institute for Astronomy,2680Woodlawn Drive,Honolulu,Hawaii 96822,USA

2School

of Physics and Astronomy,University of Birmingham,Brimingham B152TT,UK

3Department of Physics and Astronomy,Johns Hopkins University,3400North Charles Street,Baltimore,MD 21218,USA 4Joint Center for Astrophysics,University of Maryland,Baltimore County,1000Hilltop Circle,Baltimore,MD 21250,USA 5Space Telescope Science Institute,Baltimore,MD 21218,USA

6Laboratory for High Energy Astrophysics,Code 660,NASA/GSFC,Greenbelt,MD 20771,USA 7University of Maryland,College Park,MD 20742-2421,USA

8Visiting Astronomer at the W.M.Keck Observatory,jointly operated by the California Institute of Technology and the University of California.

1

2

quali?ers.Previous and ongoing cluster surveys have thus adopted either of two fundamentally di?erent,but com-plementary,approaches.For instance,very deep X-ray and optical surveys covering only small areas of sky have become increasingly successful at?nding poor clusters at z～>1(Rosati et al.,1999;Lubin et al.,2000)while,at the other extreme,relatively shallow X-ray surveys covering very large solid angles(Ebeling,Edge&Henry,2001)are in the process of producing large samples of very massive systems at lower redshift(z～0.5).

2.THE WIDE ANGLE ROSAT POINTED SURVEY

The Wide Angle ROSAT Pointed Survey(WARPS)is one of a few surveys straddling the dividing line between the two strategies outlined in the previous section.Fol-lowing the approach pioneered by the EINSTEIN Ex-tended Medium Sensitivity Survey(EMSS,Gioia et al., 1990;Stocke et al.,1991),WARPS searches for distant clusters among a large number of X-ray sources serendip-itously detected in pointed ROSAT PSPC observations. The WARPS strategy,as well as earlier results,have been discussed in previous papers(Scharf et al.,1997;Jones et al.,1998;Ebeling et al.,2000;Fairley et al.,2000).

3.DISCOVERY OF CLJ1226.9+3332

ClJ1226.9+3332was detected in the WARPS survey as an extended X-ray source14.5arcminutes o?axis in the ROSAT PSPC observation of NGC4395,a nearby,low-luminosity Sy1galaxy(RP600277N00,on-axis exposure time9036s).The signi?cance of detection exceeds10σin a2arcmin(radius)aperture.The source is also detected at more than5σsigni?cance(within an aperture of1’radius and at an o?-axis angle in14.1’)in a ROSAT HRI observation of the same target(RH702725N00,on-axis ex-posure time11,353s).

Figure1shows an I band image of the source with adap-tively smoothed X-ray contours from the ROSAT PSPC and HRI observations overlaid.The emission is well cen-tered on an apparent overdensity of faint galaxies.

3.1.X-ray observations

The net PSPC count rate directly detected by the VTP algorithm(Ebeling&Wiedenmann,1993;see also Scharf et al.1997for details of how WARPS employs this de-tection algorithm)is(2.29±0.18)×10?2ct s?1in PHA channels50to200.The exposure time at the location of the source(α=12h26m58.1s,δ=+33?32′50.1′′,J2000) is7.9ks.Assuming a beta model(Cavaliere&Fusco-Femiano,1976)withβ=2/3and core radius70h?150kpc, derived from the distribution of the immediately detected photons(see Scharf et al.1997for details),and taking into account the e?ects of the PSPC o?-axis point-spread function(PSF)we extrapolate to a total source count rate of(2.77±0.21)×10?2ct s?https://www.360docs.net/doc/02451376.html,ing the Galactic value of1.38×1020cm?2for the equivalent column density of neutral hydrogen in the direction of the source(Dickey& Lockman,1990),a metallicity of0.3,and a gas temper-ature of kT=11.7keV(see below)we convert the total PSPC count rate into a total X-ray?ux in the0.5–2.0keV band of(3.4±0.3)×10?13erg s?1cm?2.

Following the same procedure for the HRI data,we?nd an exposure time of10.2ks at the location of the source (α=12h26m58.0s,δ=+33?32′54.1′′,J2000).This source centroid agrees to within4.2”(less than the typical ROSAT astrometry error)with the one determined from the PSPC data.The extrapolated total cluster count rate is measured to be(1.10±0.18)×10?2ct s?1.We use the same assumptions as before to convert this count rate to a total X-ray?ux of(2.9±0.2)×10?13erg s?1cm?2 (0.5–2.0keV),a value that is in good agreement with the PSPC measurement.Although the HRI image shows no obvious point sources within three arcminutes of the source (see Fig.1)we cannot strictly rule out the possibility that some or all of the observed X-ray?ux originates from a concentration of individually faint point sources.

3.2.Optical observations

We obtained redshifts of11objects close to the appar-ent cluster core shown in Fig.1using the LRIS spectro-graph(Oke et al.1995)on the Keck-II10m telescope. Both apparent cluster members and possible X-ray con-taminants(blue galaxies and stars)were https://www.360docs.net/doc/02451376.html,-bining the results of a longslit observation carried out in January1999(300/5000grism,dispersion2.6?A/pixel, slitwidth1.5arcsec,wavelength range5000–10000?A)with those of multi-object spectroscopy performed in January 2000(600/7500grism,dispersion1.3?A/pixel,slitwidth1.5 arcsec,wavelength range typically6200–8700?A)we?nd no obvious contaminants(broad line emitters)but six galax-ies with accordant redshifts around a(heliocentric)mean of z=0.8877.The six spectroscopically con?rmed cluster members are marked in Fig.1;their positions and redshifts are listed in Table1.The position of the brightest clus-ter galaxy(labeled A in Table1)coincides within2”(6”) with the X-ray emission centroid as determined from the PSPC(HRI)observations discussed in Section3.1,sup-porting our identi?cation of this source as a distant cluster of galaxies.

4.INTRINSIC PROPERTIES OF CLJ1226.9+3332 From the six redshifts measured by us so far(see Ta-ble1)we obtain a very crude?rst estimate of the cluster

galaxy velocity dispersion ofσ=1650+930

?340

km/s.If con-?rmed in future observations,this very high value suggests a very massive cluster and/or substructure along the line of sight.

Assuming a mean cluster redshift of z=0.8877we de-rive a total X-ray luminosity of L X=(1.06±0.08)×1045 h?250erg s?1(0.5?2.0keV,q0=0.5is assumed throughout this Letter)from the total cluster?ux as determined from the PSPC observation,corresponding to an estimated gas temperature of11.7keV based on the L X?kT relation of White,Jones&Forman(1997).The inferred bolomet-

ric luminosity of this sytem is8.0+2.2

?1.9

×1045h?250erg s?1 (we assume an uncertainty of±1keV in the estimated kT value),making ClJ1226.9+3332more X-ray luminous than any other cluster currently known at z>0.55.The cluster’s estimated X-ray temperature is consistent with its high velocity dispersion(Mushotzky&Scharf,1997). While the X-ray and optical properties of ClJ1226.9+3332described above strongly suggest the presence of a highly massive cluster,the limited depth of the existing X-ray and optical data does not allow us to rule out contamination from a multitude of X-ray

3

faint point sources and/or from projection e?ects.Un-ambiguous evidence of the high mass of this system was, however,obtained recently in the form of the detection of a very strong Sunyaev-Zeldovich E?ect(SZE)decrement centered on the X-ray position reported here.The SZE measurements were obtained interferometrically at the Berkeley-Illinois-Maryland Array(BIMA)at a frequency of28.5GHz,and are reported by Joy et al.(2001).The BIMA observations con?rm the presence of a deep gravi-tational potential well and lead to an estimate for the total gravitational mass of ClJ1226.9+3332that is similar to, and possibly higher than that of the well studied,massive cluster MS1054.4-0321at z=0.83(Hoekstra,Franx& Kuijken2000,and references therein).

5.THE EFFECTIVE DEPTH OF SERENDIPITOUS X-RAY

CLUSTER SURVEYS

All serendipitous X-ray cluster surveys past and present, from the EMSS to the ROSAT Deep Cluster Survey (RDCS,Rosati et al.,1998),probe deep enough to detect very X-ray luminous clusters(L X>8×1044h?250erg s?1 ,0.5–2.0keV)out to redshifts z～>1.5.WARPS,for in-stance,would have detected ClJ1226.9+3332over the full geometric solid angle covered(72.0deg2)out to a maximal redshift of z=1.87.

Although formally correct,such maximal detection red-shifts are misleading.WARPS,like all other serendipitous cluster surveys conducted to date(with the notable ex-ception of the RDCS),performed its imaging follow-up observations exclusively in the optical R and I bands.For clusters at z>1,R and I correspond to B or even U in the cluster rest frame and are thus very ine?cient bands for the detection of distant cluster ellipticals.Indeed,and not surprisingly,WARPS9failed to detect any cluster at z>1.We thus argue that the limitations of the optical imaging observations impose a more stringent constraint of z max～1.

Imaging of X-ray selected cluster candidates at near-infrared wavelengths(as conducted by the RDCS team) allows the identi?cation of clusters beyond z～1(Rosati et al.,1999).However,instrumental limitations still im-pose a redshift limit of z～1.4as,at z>1.4,the most important absorption features(the Ca H and K doublet) are redshifted to observed wavelengths redward of9,500?A where the e?ciency of even the most powerful present-day spectrograph(LRIS on the Keck10m telescope)is less than5per cent.Consequently even the currently deepest X-ray cluster survey,the RDCS,has yet to spectroscopi-cally con?rm a cluster at z>1.3.

6.THE EVOLUTION OF THE MOST MASSIVE CLUSTERS

OF GALAXIES OUT TO Z～1

We attempt to constrain the evolution of the X-ray cluster luminosity function(XLF)at the highest redshifts and highest luminosities using the small,but statistically complete WARPS sample de?ned by the selection crite-ria z>0.8and L X>5×1044h?250erg s?1(0.5?2.0 keV).The only two WARPS clusters meeting these crite-ria,ClJ0152.7–1357(Della Ceca et al.,2000;Ebeling et al., 2000)and ClJ1226.9+3332,are the most X-ray luminous distant clusters currently known10.Both ClJ0152.7–1357 (which was missed by the EMSS,Ebeling et al.,2000) and ClJ1226.9+3332feature X-ray?uxes well above the WARPS?ux limit and would have been detected out to maximal redshifts of z=1.57and z=1.87,respectively. However,for the reasons described in the previous section, we use a lower maximal redshift of z=1when computing the WARPS search volumes.

The resulting cumulative WARPS X-ray luminosity function of very X-ray luminous(L X>5×1044h?250erg s?1,0.5?2.0keV)and very distant(z>0.8)clusters is shown in Figure2.Standard1σerrors(in the Poisson limit,Gehrels1986)are indicated by the dark,shaded re-gion.We also compute an alternative version of the XLF, using the very generous assumption of z max=1.4(shown by the hatched region in Fig.2).The currently best de-termination of the local(z<0.3)cluster XLF from the ROSAT Brightest Cluster Sample(Ebeling et al.,1997) is shown by the thick,solid line.We?nd both WARPS XLF estimates to be in very good agreement with the lo-cal measurement.Since the search volumes of WARPS and EMSS for clusters with L X～9×1044h?250erg s?1(0.5?2.0 keV)at0.8

Also shown in Fig.2is the prediction of the XLF evo-lution model derived by Rosati et al.(2000)from a?t to the RDCS data,together with its2σerrror range.At z～0.9this model lies between a factor of100and a factor of1000below the WARPS XLF measurement.While our data thus do not support the RDCS model,they cannot strictly rule it out either because of the large uncertain-ties in the model parameters.We note though that at L X=5?11×1044h?250erg s?1(0.5?2.0keV)the ob-served WARPS XLF for very X-ray luminous clusters at 0.8≤z≤1is inconsistent with the RDCS prediction at greater than2σsigni?cance.

7.CONCLUSIONS

The discovery of ClJ1226.9+3332,the most X-ray lumi-nous distant cluster currently known,adds to the growing evidence in favour of an early period of cluster formation at redshifts z～>1with little evolution in the cluster abun-dance ever since.

While the small size of the WARPS subsample discussed here does not allow us to?rmly rule out negative or pos-itive evolution at z～1,we stress that our data certainly do not require a change in the abundance of very X-ray luminous clusters out to the highest redshifts probed by current surveys.Speci?cally,we do not?nd a decrease in the comoving cluster abundance by a factor of more than 100(at z～0.9compared to the local value)as predicted by the best-?tting XLF model derived from RDCS data (Rosati et al.,2000).

The lack of signi?cant evolution observed by us at the highest redshifts and highest X-ray luminosities is consis-tent with the EMSS results at z>0.8if the known incom-pleteness of the EMSS at these redshifts is corrected for.

9as well as all other cluster surveys relying exclusively on imaging observations in the optical passband for cluster identi?cations 10A discussion of the W ARPS XLF of less luminous systems at z>0.8will be presented elsewhere.

4

Our?ndings also agree with preliminary results from the MAssive Cluster Survey(MACS,Ebeling,Edge&Henry, 2000)which,boasting greatly improved statistics,?nds only very mild negative evolution in a measurement of the XLF of similarly X-ray luminous clusters at z～0.4. Unless cluster evolution is a non-monotonic function of redshift and/or X-ray luminosity,our results are in con-?ict with earlier claims of highly signi?cant(>3σ)nega-tive evolution at lower redshifts(0.3≤z≤0.6)and lower luminosities(L X～>3×1044h?250erg s?1,0.5?2.0keV) based on the cluster samples compiled in the EMSS(Gioia et al.,1990)and the CfA160deg survey(Vikhlinin et al., 1998a).

Larger,well de?ned samples of massive clusters at inter-mediate and high redshift as well as spatially resolved clus-ter temperatures are needed to actually measure the rate of evolution and to place meaningful constraints on the cosmological parameters governing structure formation.

We thank the telescope time allocation committee of the University of Hawai‘i for their generous support of the WARPS optical follow-up programme.HE gratefully acknowledges?nancial support from NASA LTSA grant NAG5-8253.LRJ thanks the UK PPARC for?nancial support.

REFERENCES

Cavaliere,A.&Fusco-Femiano,R.1976,A&A,49,137

Della Ceca R.,Scaramella R.,Gioia I.M.,Rosati P.,Fiore F.,Squires

G.2000,A&A,353,498

Dickey,J.M.&Lockman,F.J.1990,Ann.Rev.Astron.Astroph.,28, 215

Ebeling H.,Edge A.C.,Fabian A.C.,Allen S.W.,Crawford C.S., B¨o hringer H.1997,ApJ,479,L101

Ebeling H.,Edge A.C.&Henry J.P.2000,in Large-Scale Structure in the X-ray Universe,eds Plionis&Kolokotronis,Atlantisciences, p39

Ebeling H.,Edge A.C.&Henry J.P.2001,ApJ,submitted Ebeling H.et al.2000,ApJ,534,133

Fairley B.W.,Jones L.R.,Scharf C.,Ebeling H.,Perlman E.,Horner

D.,Wegner G.,Malkan M.2000,MNRAS,318,333

Gehrels N.1986,ApJ,303,336

Gioia I.M.,Maccacaro T.,Schild R.E.,Wolter A.,Stocke J.T.,Morris S.L.,Henry J.P.1990,ApJS,72,567

Gioia I.M.,Henry J.P.,Maccacaro T.,Morris S.L.,Stocke J.T., Wolter A.1990,ApJ,356,L35

Hoekstra,H.,Franx,M.,Kuijken,K.2000,ApJ,532,88

Jones L.R.,Scharf C.,Ebeling H.,Perlman E.,Wegner G.,Malkan M.,Horner D.1998,ApJ,495,100Joy M.et al.2001,ApJL,submitted,astro-ph/0012052

Lubin L.M.Brunner R.,Metzger M.R.,Postman M.,Oke J.B.2000, ApJ,531,L5

Mushotzky R.F.&Scharf C.A.1997,ApJ,482,L13

Oukbir J.&Blanchard A.1997,A&A,317,10

Rosati P.,Della Ceca R.,Norman C.,Giacconi R.1998,ApJ,492, L21

Rosati P.,Stanford S.A.,Eisenhardt P.R.,Elston R.,Spinrad H., Stern D.,Dey A.1999,AJ,118,76

Rosati P.,Borgani S.,Della Ceca R.,Stanford S.A.,Eisenhardt P.R., Lidman C.2000,in Large-Scale Structure in the X-ray Universe, eds Plionis&Kolokotronis,Atlantisciences,p13

Scharf C.A.,Jones L.R.,Ebeling H.,Perlman E.,Malkan M.,Wegner

G.1997,ApJ,477,79,Paper I

Stocke J.T.,Morris S.L.,Gioia I.M.,Maccacaro T.,Schild R.,Wolter

A.1991,ApJS,76,813

Viana P.T.P.&Liddle A.R.1996,MNRAS,281,323

Vikhlinin A.,McNamara B.R.,Forman W.,Jones C.,Quintana H., Hornstrup A.,1998a,ApJ,498,L21

Vikhlinin A.,McNamara B.R.,Forman W.,Jones C.,Quintana H., Hornstrup A.,1998b,ApJ,502,558

White D.A.,Jones C.&Forman W.1997,MNRAS,292,419

Table1

Positions and heliocentric redshifts of confirmed cluster members galaxy R.A.(J2000)Dec(J2000)redshift z

Note.—The quoted positions are accurate to better than

1arcsec.All redshifts were derived from spectra taken with

LRIS on Keck-II.The quoted redshift errors are1σstandard

deviations from the quoted mean and are based on the ob-

served wavelengths of individual spectral features.

5

12552705A B

C

D

E

F

h 26m 50s

s m 00s

s

Right Ascension (2000)

33° 31′ 30″

32′ 00″

30″

33′ 00″

30″

34′ 00″

30″

D e c l i n a t i o n (2000)

Fig. 1.—I band image of ClJ1226.9+3332obtained in a four minute exposure with LRIS on Keck-II in February 1999.Overlaid are logarithmically spaced contours of the adaptively smoothed X-ray emission as seen with the ROSAT PSPC in the 0.5–2.0keV band (solid lines),and as seen with the ROSAT HRI (dashed lines).The lowest PSPC (HRI)contour lies a factor of 2.5(1.2)above the background value of 2.5×10?4(5.4×10?3)ct s ?1arcmin ?2;the levels of adjacent contours di?er by a factor of 1.5(1.2).The FWHM of the PSPC (HRI)point-spread function at this o?-axis angle (14arcmin)is 37(6)arcsec.

6

10

10101010

L X (1044

h 50

?2

erg s ?1

, 0.5 ? 2.0 keV)

n (>L X ) (h 503 M p c ?3)

5

Fig.2.—The cumulative X-ray luminosity function (XLF)of very X-ray luminous,distant (z >0.8)clusters based on ClJ0152.7–1357and ClJ1226.9+3332as detected in the W ARPS survey (solid lines).The strongly shaded region shows the 1σPoisson error range around the XLF assuming a maximal detection redshift of z =1;the hatched region just below delineates the 1σPoisson error range for the W ARPS XLF assuming a maximal detection redshift of z =1.4.The local XLF based on the ROSAT Brightest Cluster Sample (Ebeling et al.,1997)is shown by the bold,solid line,whereas the prediction of the RDCS XLF evolution model (Rosati et al.2000)with its 2σerror range is shown by the dashed line and the lightly shaded region.